An incentive system for decentralized DAG-based platforms

Abstract

Decentralized public platforms are becoming increasingly popular due to a growing number of applications for various areas of
business, finance, and social life. Authorless nodes can easily join such networks without any confirmation, making a transparent
system of rewards and punishments crucial for the self-sustainability of public platforms. To achieve this, a system for incentivizing
and punishing Workers’ behavior should be tightly harmonized with the corresponding consensus protocol, taking into account all of
its features, and facilitating a favorable and supportive environment with equal rights for all participants. The main purpose of rewards is to incentivize Workers to follow the protocol properly, and to penalize them for any type of misbehavior. The issues of
block rewarding and punishing in decentralized networks have been well studied, but the DAG referential structure of the distributed
ledger forces us to design methods that are more relevant. Since referential structures cannot be reliably validated due to the fact that
they are built on the basis of the instantaneous visibility of blocks by a certain node, we propose to set rewards for blocks in the DAG
network based on the degree of confidence of topological structures. In doing so, all honest nodes make common decisions based
only on information recorded into the ledger, without overloading the network with additional interactions, since such data are always identical and available. The main goal of this work is to design a fair distribution of rewards among honest Workers and establish values for penalties for faulty ones, to ensure the general economic equilibrium of the Waterfall platform. The proposed approach has a flexible and transparent architecture that allows for its use for a wide range of PoS-based consensus protocols. The core
principles are that Workers' rewards depend on the importance of the conducted work for block producing and achieving consensus
and their penalties must not be less than the potential profit from possible attacks. The incentivizing system can facilitate protection
from various kinds of attacks, namely, so-called Nothing-at-stake, Rich-get-richer, Sybil, and Splitting attacks, and from some specific threats related to a DAG structure.